ARTICLE
Distributive Justice in National Security Law
Daphne Barak-Erez*
Abstract
When collecting information about possible terrorist attacks, national
security agencies may have to choose between competing systems of
implementation, all infringing individual rights. Should they collect
information by indiscriminately wiretapping communications in the
population at large or by implementing harsher means, such as
investigations under arrest, against individuals known to be involved in
terrorist activities? Assuming that both policy options, at least in certain
cases, are equally effective, this question highlights the fact that many
national security decisions entail distributive implications. This Article
analyzes the centrality of distributive justice considerations for the
understanding of national security law at a time that policy decisions on
anti-terrorism measures can either take the form of indiscriminate largescale infringements of individual rights, such as body-scans, or much more
targeted forms, which place additional burdens on suspects. This choice,
which had so far been overlooked, becomes more and more relevant as new
technologies make it easier to apply methods of surveillance
indiscriminately.
The focus on the choice between policies with different distributive
implications cuts against the traditional understanding of national security
law as based on two choices: allowing national security threats to continue
unabated, which may result in a catastrophe ensuing from a terrorist attack,
or violating rights through the government's use of preventive measures
*Justice.,Supreme Court of Israel. Formerly Stewart andJudy Colton Professor of Law and
the Chair of Law and Security., Faculty of Law., Tel Aviv University. I thank the
participants of the faculty workshops at Duke Law School., Virginia Law School and the
Interdisciplinary Center, as well as Avichai Dorfman, David Enoch, Talia Fisher, Shimrit
Itay., Roy Kreitner, Galit Raguan, Yuval Roitman, Paul Sass, Paul Stephen, Matt
Waxman, and Ernie Weinrib for their helpful comments, and Ofra Bloch and Naomi
Scheinerman for their research assistance.
Copyright C 2012 by the Presidents and Fellows of Harvard College and Daphne BarakErez
2012 /Distributihe
Justice in NationalSecurity Law
284
designed to confront the threat. This pattern of analysis has obscured a
fundamental question to the implementation of measures for preventing
anti-terrorism: what is ajust distribution of the resulting burdens?
This Article examines the hidden side of national security law by focusing on
the choice between using harsh anti-terrorism measures on a selective basis,
or using anti-terrorism measures that entail less harmful infringements of
individual rights but affect much larger segments of the general population.
Introduction: The Many Faces of Distribution in National Security Law
Faced with an ever-pressing need for collecting information about
possible terrorist attacks, what course should national security agencies
prioritize? Should they collect information by indiscriminately wiretapping
communications in the population at large, or concentrate their energies on
investigating those known to be involved in terrorist activities, using harsher
means such as investigations under arrest? Each of these alternatives
allocates the cost of national security to a different group. Assuming that
both policies are (at least sometimes) equally effective, this choice and others
described in this Article highlight the fact that national security decisions
often entail distributive implications that need to be taken into
consideration. Even the decision to refrain from anti-terrorism measures is
laden with hidden distributive consequences: it allocates the costs of being
hurt to individuals more likely to be attacked, such as residents of large
metropolitan areas, and prioritizes the interests of those who may have
suffered from anti-terrorism measures.1 In the service of promoting security,
governments constantly choose how to allocate their costs among different
groups. Given that these hidden decisions underlie the national security
policies of almost every government in the world, the significance of
recognizing the role that distributive justice concerns should play in the
formulation of these policies can hardly be exaggerated.
The choices available in the area of national security become more
pressing due to the development of new technologies. These technologies
have opened up new possibilities for the use of large-scale anti-terrorism
measures. One controversial example is the use of body-scans in airports on
a universal basis as an alternative to the more selective procedures used in
I For a comparison of the costs of action and inaction in the context of preventive antiterrorism measures, see ALAN M. DLRSHOWITZ, PRLLMPTION: A KNIFE THAT CUTS BOTH
WAYS 245 (2006).
285
HarvardNational Security Journal/ Vol. 3
the past.2 In light of these developments, the need to consider the
distributive implications of each governmental choice in this area becomes
even timelier.
Attempts to analyze the distributive aspects of national security law
seem antithetical to the conventional wisdom that views national security
policies as the ultimate example of a "public good," that is, something that
all individuals in society enjoy and share the costs of equally (in the typical
case, through taxation).3 This conventional wisdom has obscured the fact
that national security decisions, especially in modern times, have multidimensional distributive consequences. Theories of distributive justice
attempt to define the correct balance of the burdens borne and the benefits
received by particular citizens or groups of people. So far, they have been
used to analyze basic concepts in several areas of law such as tort liability,4
taxation, and contracts, 6 but have never been central to the discourse on
national security law and policy. This Article seeks to broaden the
perspectives used to understand national security law by highlighting the
centrality of distributive justice in every decision fitting under the rubric of
"national security."
this controversy, see, e.g.., Charles Krauthammer, Don't Touch My Junk, WASH. POST
(Nov. 19, 2010), http: //www.washingtonpost.com/wpdyn/ content/ article/ 2010 /11/ 18 /AR2010111804494.html;Jeffrey Rosen, Why the PatDowns and Body Scans are Unconstitutional,WASH. POST (Nov. 28, 2010),
http://www.washingtonpost.com/wpdyn/ content/ article/ 2010/11/ 24/AR20101124045I10.html.
2 For
MANCUR OLSON, THE LOGIC OF COLLECTIVL ACTION: PUBLIC GOODS AND THE
THEORY O1 GROUPS 14 (1965) ("The basic and most elementary goods or services
provided by government, like defense and police protection, and the system of law and
order generally., are such that they go to everyone or practically everyone in the nation. It
would obviously not be feasible, if indeed it were possible., to deny the protection provided
by the military services, the police and the courts to those who did not voluntary pay their
share of the costs of government."); Ann R. Markusen, 7e Case Against PriVatizingNational
Sectioity 16 GOVERNANCL 471, 473 (2003) ("The nature of national security as a public good
has been understood for decades and is noncontroversial. The basic argument is that
national defense is both nonexcludable and nondepletable. There is no way of providing it
in a decentralized, 'fee for service' manner-everyone would have an incentive to take a
'free ride' on their neighbors., as., indeed, some countries do.").
See, e.g., Gregory C. Keating., Distributive and CorrectiveJustice in the Tort Law ofAccidents, 74 S.
CAL. L. REV. 193 (2000).
- See, e.g., Linda Sugin, 7ieofies ofDistfibutiveJustice and Limitations on Taxation: 1I7hat Rawls
Demandsjrom Tax Systems., 72 FORDHAM L. REV. 1991 (2004).
See, e.g., Anthony T. Kronman, Contiact Law andDistuibutieJustice,89 YALE L.J. 472 (1980
2012 /Distributihe Justice in National Security Law
286
In the national security realm, distributive justice theory considers
the allocation of both the benefits and the burdens of measures applied
invoking security. The benefits of such measures include the protection and
sanctity of life, while the burdens, in addition to monetary costs, include
infringements of individual rights such as privacy, liberty, and, at times, life.
These benefits and burdens, however, do not fall equally on all parts of the
population. For example, although protection from terrorist attacks in
airports and central public sites are perceived as necessary for ensuring the
security of the state and the public at large, pilots and frequent air travelers
directly benefit more than others from anti-terrorism measures that aim to
lessen such risks. Similarly, those who are subject to profiling in airports
carry an additional burden in the name of the promotion of a good that is
enjoyed not mainly by them, but rather by others in the public at large.
So far, the literature has acknowledged the distributive justice
aspects of national security decisions in a relatively marginal manner.
Ronald Dworkin focused on the price paid for security and argued that the
balancing approach to security and civil rights was problematic since antiterrorism measures do not diminish civil liberties for everyone. His
argument focused on the uneven allocation of the cost of security, stating
that "[m]ost of us pay almost nothing in personal freedom when such
measures are used against those the President suspects of terrorism." 7
Looking at the other side of the balance between security and rights-the
side of the benefit of national security policies Jeremy Waldron discussed
the uneven distribution of security as a "good." 8 He pointed out that
"people may be differently situated with regard to a given threat," 9 and
described how government decisions to promote the security of one group
often jeopardize the security of another.' 0
The distributive aspects of national security policies have left
significant marks, mainly on the analysis of policies that have uneven effects
for different ethnic groups. Ethnic profiling has occasionally been addressed
from the perspective of distributive justice, pointing that this practice
7Ronald Dworkin, The Threat to Patiotism, N.Y. RLV. BOOKS, Feb. 28, 2002, at 44, 48.
SeeJeremy Waldron., Safety and Secmity 85 NEB. L. RLV. 454, 460 (2006).
Id. at 484: see also Rafael Di Tella & Ernesto Schargrodsky, Do Police Reduce Cuime? Estimates
Using the Allocation ofPolice Forces After a TerrouistAttack, AM. ECON. RLV. Mar. 2004, at 115
(discussing the benefits that accrued to residents of areas neighboring Jewish institutions,
which enjoyed additional police presence following the bombing attack on the main Jewish
Center in Buenos Aires in 1994).
1(oWaldron, supra note 8, at 485-89.
287
HarvardNational Security Journal/ Vol. 3
allocates an additional cost to a certain group in society. Some proposals
have even sought to compensate individuals who have been subjected to
questioning or other related burdens."
This Article aspires to focus on the distributive implications of
decisions in the realm of national security in a more comprehensive manner.
First, it argues that the distributive aspects of such decisions should be
recognized as a major theme rather than a marginal aspect of national
security law. Second, it looks into new questions that have not yet been
exposed and discussed-mainly into the need to choose between harsh antiterrorism measures on a selective basis and anti-terrorism measures that
entail less harmful infringements of human rights but affect large
populations of innocent individuals. Third, it moves from criticism of the
traditional balancing approach to the mapping of the substantive and
institutional considerations that should guide the analysis of choices made in
the area of national security law.
Following the Introduction, Part I describes in greater detail the
distributive choices entailed by almost every national security decision. Part
II presents national security measures as regulatory decisions and describes
their distinctive characteristics (e.g., the need to compare between different
kinds of harms, the significance attached to the moral responsibility of those
targeted by the government and their political "otherness"). Part III
examines the challenge of distributive justice in the area of national security
by looking at specific case studies. Part IV broadens the perspective of the
analysis by highlighting the relevance of institutional considerations, such as
the relative disadvantages of the courts in dealing with considerations of
distributive justice. Part V builds on the analysis in the previous sections and
offers an initial scheme for thinking about distributive justice in the area of
national security by addressing the relative advantages and disadvantages of
targeted anti-terrorism measures vis-a-vis anti-terrorism measures applied
on a universal basis. Accordingly, the conclusions stress the importance of
informing national security decisions with their distributive implications,
and shed additional light on the relevant considerations in this regard.
I1 Tomer Blumkin & Yoram Margalioth., On Terror, Dings and RacialProfiling.,8 INT'L RLV.
OF L. & ECONOMICS 194, 200 (2008). For an analysis of racial profiling, see also infia text
accompanying notes 50-52.
2012 /Distributive Justice in National Security Law
288
I. The Distributive Choices Hidden in National Security Law
How should societies allocate the burdens of providing national
security? Assuming that all possible methods are equal in aspects such as
effectiveness, financial cost, and efficiency, what should be the principle
guiding the choice of method? This classic distributive justice question of
"holding all else equal" so common in other areas of public policy has rarely
been studied in the context of national security. Traditionally, national
security law has been understood only as a choice between allowing a threat
to national security to continue unabated (e.g. the catastrophe that may
result from a terrorist attack) and infringing human rights via the
government's use of preventive measures when confronting a threat. In the
context of this classic discourse, anti-terrorism measures are considered
justified, despite their effect on human rights, if they are likely to prevent a
greater evil to society. An illuminating instance of this thinking is the
opening of Michael Ignatieffs book, The Lesser Eil: "What lesser evils may a
society commit when it believes it faces the greater evil of its own
destruction? This is one of the oldest questions in politics and one of the
hardest to answer." 12 Similarly, Eric Posner and Adrian Vermeule base
their discussion of national security law on a binary choice between
"security" and "liberty," which they describe as a "tradeoff." 3
As indicated, Dworkin had already criticized this approach as
overlooking distributive considerations.' 4 Moving from criticism to
implementation, this Article poses the following question: how should the
burdens necessary to attain the benefit of preventing anti-terrorism be
distributed among individuals and groups in society? Typically, there are
many options for ensuring prevention. Generally, anti-terrorism measures
can be grouped into measures that target only those suspected of
12 MICHALE IGNATILFF, THE LESSER EVIL: POLITICAL ETHICS IN AN AGL O TERROR 1
(2004). Gabriella Blum asks this question in the area of humanitarian law. See Gabriella
Blum, 7ie Laws of War and the 'Lesser Ezil' 35 YALLJ. INT'L L. 1 (2010). It may indeed be the
case that, in some contexts, it is possible to take anti-terrorism measures that do not affect
human rights (what Kent Roach calls "smart" strategies), and even that, in other contexts,
measures that infringe human rights are not necessarily effective (what Roach calls "harsh"
measures). See Kent Roach, Must We Trade Rightsfor Secnity? 7ie Choice between Sniart, Haish,
or Proportionate Secmity Strategies in Canada and Britain, 27 CARDOZO L. REV. 2151, 2152-54
(2006). However, even when we disregard these examples., we retain the dilemma that
necessitates a trade-off between human rights and security.
ERIC A. POSNER & ADRIAN VLRMLULE, TERROR IN THL BALANCL: SECURITY,
LIBLRTY, AND THL COURTS 21 (2007).
'3
14 Dworkin, supra note 7, at 48.
289
HarvardNational Security Journal/ Vol. 3
participating in terrorist activity (be it through harsh interrogations,
detentions, or control orders) and measures that affect the general public (be
it through the use of biometric identification tools, limitations on the
distribution of ideas advocating terrorism, random surveillance of telephone
conversations and other forms of communication, or restricting access to
sights that may serve as targets). Consider the example of collecting
intelligence information on the activities of terrorist organizations, which
can be attained through harsh means of interrogation directed at members
of such organizations, or through the relatively large-scale use of random
surveillance of phone conversations of the general population. Let us assume
that both measures will lead to similar results in terms of effectivenesswhich one is better? This dilemma may bring to mind earlier challenges to
national security. In the 1950s, for example, should the U.S. government
have abstained from hiring people associated with the Communist party,
thereby imposing a significant burden on this group, or should it have opted
for the close monitoring of all government employees?'1
Addressing the question of distributive justice in this way raises two
pragmatic concerns: first, the effectiveness of the two alternatives compared
may not be identical in real life scenarios, and second, a large cloud of
uncertainty surrounds the effectiveness assessment of national security
measures. But although these concerns are real, they do not undermine the
question posed. Normative concerns remain even when measures differ in
their effectiveness, the only difference being that they will then have to be
considered vis-A-vis the additional variable of relative effectiveness. The
overall skepticism concerning the ability to assess the effectiveness of
government initiatives in the area of national security is also important, but
not unique to the dilemma at hand. This concern, if overstated, may
undermine the legitimacy of any national security initiative, not only when it
is compared to another alternative. Developing techniques for assessing the
effectiveness of national security measures is indeed essential. This
additional challenge, however, exceeds the scope of this discussion.
The core of my argument is that national security law should not be
understood as a choice between two-dimensional extremes: less antiterrorism measures and a higher probability of an attack or more antiNeedless to say., the focus on the comparison between broadly used measures and
narrowly used measures is important., because when the narrowly used measures are of the
same severity, a choice that prioritizes the use of the narrowly used measures seems easy as
long as the narrow group is not selected on a discriminatory basis.
1)
2012 /Distributihe
Justice in NationalSecurity Law
290
terrorism measures and a lower probability of an attack. Rather, it involves
three-dimensional choices-which have to consider the damages entailed by
potential terrorist attacks, the relatively serious infringements of the rights of
a relatively small group of suspects when targeted anti-terrorism are applied,
and the relatively milder infringements of the rights of much larger segments
of the general population when large scale measures are applied. Assuming
that the greater evil of terrorism should be prevented, and that its
prevention is a benefit demanded by and for society and hence an obligation
of the government and of policy makers, the challenge is how to choose
between these options. Anti-terrorism law thus evolves into an area that
posits questions of distributive justice: who should bear its burdens and
receive its benefits, and how much of each?
The analysis below does not pretend to solve this dilemma but rather
to point out the need for assessing the considerations that should be involved
when such decisions are contemplated. In many ways, as Calabresi and
Bobbitt have noted, these decisions have the characteristics of a "tragic
choice" in that they inevitably allocate agony, suffering, and even death.16
II. A Regulatory Challenge: The Special Nature of Distributive Justice
Considerations in a National Security Context
The use of distributive justice for the purpose of allocating social
burdens in a national security context deviates significantly from this
theory's original setting. Traditionally, distributive justice concepts were
analyzed in the context of allocating goods and benefits,' 7 and not for the
IGGUIDO CALABRESI & PHILIP BOBBITT, TRAGIC CHOICLS 15 (1978). Calabresi and
Bobbitt study examples in which society is faced with the need to allocate essential scarce
resources. The need to allocate the burden of national security has the same features.
17 Aristotle, who was the first to define distributive justice (and distinguish it from corrective
justice), stated that it is "manifested in distributions of honour or money or the other things
that fall to be divided among those who have a share in the constitution." ARISTOTLE, V
NICOMACHLAN ETHICS 2 (W.D. Ross, trans.) (350 B.C.), available at
http://classics.mit.edu/Aristotle/nicomachaen.5.v.html. The relatively modern notion of
distributive justice began to receive attention with the works of Adam Smith at the start of
the industrial revolution in eighteenth-century England. Factory workers lived in abject
poverty and disease with minimal standards of living and confronted a stringent social
hierarchy that made mobility near impossible, while their productivity made possible an
unprecedented surplus of goods. The application of concepts ofjustice to the condition of
poverty produced the modern theories of distributive justice. Authors and philosophers
were charged with the task of determining how these goods should be distributed in society,
291
HarvardNational Security Journal/ Vol. 3
purpose of allocating costs and burdens.' 8 But benefits and burdens are
almost inherently interconnected, and distributive justice concepts have long
been discussed in the context of allocating burdens in several other areas of
law.19 Moreover, distributive justice concerns have also been used when
considering the moral justification of self-defense, a moral query clearly
relevant to the area of national security law. One argument adduced in this
area has been that self-defense shifts the burden of the potential injury from
innocent persons to their attackers and, therefore, from a distributive
perspective, implies a better allocation of the burden. 20
An analysis of the distributive justice implications of national security
decisions must be based on understanding them as regulatory decisions.
Decision makers are expected to regulate the risks to national security in a
manner that will address the need for choosing between the allocation of
heavy burdens to a small group of people and the allocation of lesser
burdens to larger numbers of people. This is not entirely unique, and
regulatory choices involving distributive components are likely to exist in
other contexts of regulation. 2' I will argue, however, that they have some
special features in the area of national security.
The considerations characterizing regulatory choices in the area of
national security law can be summarized in the following four points: First,
the burdens to be balanced in a national security context differ in kind,
comprising, for example, on the one hand, the deprivation of liberty of
suspects in terrorist activity and, on the other, the inconvenience of people
checked at airports. Second, a moral responsibility factor is often present in
from the lowest employee to the highest manager. See SAMUL FLEISHACKER, A SHORT
HISTORY OF DISTRIBUTIVLJUSTICL 62 (2004).
18SeeJOHN RAWLS., A THLORY OFJUSTICE 25 (1971).
1BSee supra notes 4-6.
2o Phillip Montague., Se//Defense and ChoosingAmong Lives, 40 PHIL. STUD. 207, 215-16
(1981); Michael Moore, Torture and the Balance ofEills, 23 ISR. L. REV. 280, 283-86 (1989).
For a criticism of these views, see Whitley Kaufman, Tortuie and the 'DistldbutiveJustice' Heoy
of SelfDefense: An Assessment, 22 ETHICS AND INT'L AFF. 93, 100-12 (2008). For distinctions
in self-defense law between attackers who are culpable actors and innocent people who
constitute threats to others, see Shlomit Wallerstein, Justfyng the Right to Se//Defense: A 7 lleor
ofForced Crcunstances, 91 VA. L. REv. 999, 1001-02 (2005).
21 See, e.g., CASS SUNSTLIN, THL COST-BENEFIT STATE: THL FUTURL O RLGULATORY
PROTLCTION 82 (2002) ("In addition to knowing the benefits and costs of regulation, it is
necessary to know who bears those costs and enjoys those benefits, and also the particular
nature of those costs and benefits") (emphasis in original).
2012 /Distributihe
Justice in NationalSecurity Law
292
the sense that the people targeted for harm by decision makers may, in
certain cases, also be those responsible for the threat to national security.
Third, decisions in the area of national security quite regularly involve
considerations of political affiliation and participation because attackers are
often foreigners or members of distinct communities who do not participate
fully in the life of general society or are poorly represented in its politics and
institutions. These communities are therefore less likely to be harmed by
terrorist attacks on the one hand, but have a greater chance of being
targeted for harm by decision-makers who shape anti-terrorism policies on
the other. Fourth, the burdens allocated by decision-makers are generally
costs that cannot be easily spread to others and, therefore, the government's
initial choice of evils will probably not allow for cost sharing. The relevance
of these unique features of distributive justice and their effect on national
security decisions are discussed below in greater detail.
The first feature of national security decisions is that they require a
comparison between alternatives that are not only quantitatively but also
qualitatively different. This is the case, for example, when decision makers
must choose between relatively harmful and more intrusive measures
affecting a small group of people suspected of terrorist involvement, and
relatively less intrusive, large-scale measures that affect the general
population. This comparison is obviously very problematic. From a
deontological perspective, it may even seem impossible, given the Kantian
categorical imperative to avoid wrongdoing even when it appears to lead to
good results.2 2 In fact, a choice between widely infringing the rights of a
small group of suspected terrorists and slightly infringing the rights of a large
group of people is not necessarily easy even from a utilitarian perspective.
Indeed, utilitarians use the concept of aggregate harm, derived from the
concept of aggregate happiness,23 when considering regulatory alternatives
22 This approach is best exemplified in Kant's "on a Supposed Right to Lie from
Philanthropy." SeeJacob Weinrib., Thejuuidical Signficance of Kant's 'Supposed Rigit to Lie', 13
KANTIAN RL\IL\\ 141, 141-42 (2008). It is worth noting., however, that some interpret
Kant's view as applying to parties who are "autonomous and committed to reciprocity, at
least in a minimal sense." Tamar Schapiro., Kantian Rigorism and Mitigating Circunstances, 117
ETHICS 32, 56 (2006). According to this view, some acts that are usually regarded as
prohibited under Kantian law may therefore, given this caveat, be possible vis-a-vis
terrorists who act in a manner that rejects any form of reciprocity or consideration of its
victims' right to live.
2
'JERLMY BLNTHAM, INTRODUCTION TO THE PRINCIPLLS OF MORALS AND LEGISLATION
121-23 (Batoche Books, 2000) (1781); see also BENJAMIN CONSTANT, PRINCIPLES OF
POLITICS APPLICABLE TO ALL GoV ERNMENTS (1815).
HarvardNational Security Journal/ Vol. 3
293
and seeking solutions that involve a lesser aggregation of harms and thus
maximize aggregate happiness.2 4 Even they, however, usually apply this
form of argumentation when all the harms involved are equally or almost
equally severe, implying only small inequalities in harm. A utilitarian
response that incorporates the aggregation approach would therefore opt
for harming a smaller group, but the analysis changes when the harms
compared are not of equal severity. Under these circumstances, even harm
to a very large group may be a lesser aggregation of harms when the harm is
different in kind.2 5 According to John Stuart Mill, the goal of society is
indeed happiness, but what is morally right cannot be defined merely by
what maximizes happiness. Rather, an act is morally right and hence
obligatory if society can compel an individual to perform that act.
Therefore, who derives happiness and pleasure from an act is of paramount
importance to Mill, and the universal aggregate acquisition of happiness,
possible in greater or smaller measure, is not necessarily the goal. 26
The second feature of national security decisions concerns the
responsibility for the threat that the government chooses to face through the
24 For an in depth analysis of the aggregation approach, see T. M.
SCANLON, WHAT WL
OWL TO EACH OTHER (1998); Sophia Reibetanz, Contractualismand Aggregation, 108 ETHICS
296 (1998).
25 Scanlon writes
... contractualism supports a principle according to which, in situations in
which aid is required and in which one must choose between aiding a
larger or a smaller number of people all of whom face harms of
comparable moral importance, one must aid the larger number. On the
other hand, contractualism does not require., or even permit, one to save
a larger number of people from minor harms rather than a smaller
number who face much more serious injuries. This distinction, between
one class of cases, in which the number of people who can be saved is
morally relevant, and all others, in which it is not, is subject to at least
two objections. The first concerns the way in which a distinction is drawn
between the moral significance of different harms. It seems implausible
that in one case, in which we must choose between saving one person
and saving ten from harms of the same degree of seriousness, we are
required to save the ten, but that in a case that was otherwise identical
except for the fact that the harm faced by the one was slightly worse we
would be required to save the one instead. The proper reply here, I
believe, is that the distinctions on which the principles I have argued for
rely are distinctions between broad categories of moral seriousness.
Scanlon, supra note 24., at 238.
lLJOHN STUART MILL., UTILITARIANISM 72-75 (Univ. Chicago Press, 1906) (1861).
2012 /Distributihe Justice in National Security Law
294
implementation of harmful measures. Here, then, the state confronts a
threat posed by individuals who are autonomous actors and can be held
responsible for their actions, in contrast to circumstances in which the threat
is created by a natural disaster, for example.2 7 Thus, the debate here is
similar to the distributive justice rationale mentioned in the context of selfdefense, which concerns the justice of re-allocating harm to the individual or
individuals responsible for the attack.2 8 This factor is highly relevant in the
national security arena because individuals involved in terrorist activities are
usually candidates for incurring the costs of the more burdensome antiterrorism measures (leaving aside, for the time being, issues of false
identification involving people mistakenly suspected of terrorist activity). 29
Therefore, using anti-terrorism measures aimed directly against them would
appear fairer.3 0 The concept of responsibility for the consequences of one's
actions actually demands that those guilty of terrorist activity, or of any
activity that helps to promote the aims and goals of terrorism, be subjected
to the consequences of their actions. From this perspective, then, the extent
of engagement in terrorism should translate into the extent of the national
security burdens borne by individuals who are involved in it.31
27The analysis focuses on responsibility for autonomous choices, assuming that even people
who have suffered disempowerment and deprivation are responsible for their choices when
they decide to take part in violence against innocent civilians.
28 See stpra note 20. The focus on responsibility and choice was mentioned also by Scanlon,
who wrote, "It at least seems that when a person could have avoided a certain result by
choosing appropriately, this fact weakens her grounds for rejecting a principle that would
make her bear the burden of that result." Scanlon, stpra note 24, at 256.
2' The assumption here is that it is possible to ascertain this aspect based on serious
intelligence material, although this is not always the case.
so In fact, when such anti-terrorism measures are directed against individuals who are not
directly involved in terrorism, this may serve as a basis for bringing a legal challenge against
them. This was the case in HCJ 7052/03 Adalah Legal Center for the Rights of the Arab
Minority in Israel v. Minister of Interior 61(2) IsrLR 202 [2006] (Isr.), which dealt with a
sweeping policy against immigration to Israel of Palestinians from the occupied territories
(based on a general assumption regarding the hostility of their population). The petitioners
challenged this policy and asked the Court to limit the non-immigration rule only to people
involved in anti-Israeli terrorist activity. The Israeli Supreme Court acknowledged the
harsh implications of the policy on Arab Israeli citizens who have relatives in the territories,
but the majorityjustices declined to intervene due to the temporary nature of the policy
(presented as deriving from the security conditions and not reflecting a principle planned to
apply on a long-term basis). See also Daphne Barak-Erez, Citizenship and Inmmigration Law in the
Vise ofSecurity, Nationality, and Human Rigits, 6 INT'LJ. CONST. L. 184 (2008). The Israeli
Supreme Court was split in this matter again in a later decision. See HCJ 466/07 Galon v.
The Attorney General (decision from 1 IJanuary 2012) (Isr.).
iIAn extreme version of this idea has been expressed by Herbert Spencer. See HLRBLRT
SPLNCLR, 2 PRINCIPLES O1 ETHICS § 454 (1891). Spencer's argument was raised in
295
HarvardNational Security Journal/ Vol. 3
The need to take responsibility into account in the distributive
justice equation is not unique to national security decisions. It may also be a
relevant factor in regulatory decisions that allocate the cost of
environmental risks between polluters and the general public, or in decision
making processes considering whether to burden tobacco and alcohol
manufacturers with specific taxes, as well as in other contexts. For instance,
an agency whose mandate is to lower the number of car accidents may have
to choose between revoking the drivers' licenses of people with a history of
dangerous driving or placing greater speed limits on all drivers. However,
this "responsibility" factor is not always present in other regulatory contexts
such as, for instance, the regulation of a risk of spread of an infectious
disease, when decision makers have to choose between a large-scale
vaccination campaign that will exact a price from every single individual
and the imposition of a quarantine on individuals who are infected or
suspected of being infected. Daniel Markovits argued for the preference of a
large-scale vaccination policy relying, inter alia, on the lack of moral blame
on the part of those who may be targeted for quarantine.3 2 By contrast,
national security decisions are special in that responsibility is inherent in the
events they handle.
The "responsibility" factor may be deemed problematic in the area
of national security law because of the uncertainty attached to the suspicion
regarding one's involvement in terrorist activity. Harsh measures against a
terrorist who has admitted involvement in a planned attack differ from the
use of such measures against someone suspected of wrongdoing on the basis
of intelligence material but without certainty as to his involvement. This
conjunction with his objections to the poor laws in nineteenth-century England. Spencer's
objections to the poor laws were that both the undeserving (those who did not work hard)
and the deserving (those who worked hard) would receive the same benefit, and that this
benefit would come at the unjust burden of the middle and upper classes. However, taking
the responsibility factor into account does not necessarily mean that this is the only factor
that should be given weight, neither does it mean accepting it without acknowledging the
hardships confronted by individuals whose life opportunities were limited. It is clear that
considerations of personal responsibility are not the only relevant ones. The dilemma of
weighing individual responsibility vis-At-vis other values also arises in the context of other
decisions with distributive justice implications, such as deciding on the parameters of
medical insurance coverage for people responsible for neglecting their health. See Lisa Day,
DistributiveJustice and PersonalResponsibilityfor Choices about Health, 15 AM.J. CRITICAL CARE
96, 97 (2006).
2 Daniel Markovits, Quarantines and DistributivejusticeJ.L. MLD. & ETHICS 323, 333-36
(2005).
2012 /Distributihe
Justice in NationalSecurity Law
296
factor should of course be taken into consideration, as explained below.
Once national security is understood as a matter of regulating risks,
however, an action based on the evidence gathered by the government is
inevitable. In fact, this is the standard practice when agencies regulate
environmental or health risks without knowing their causes for certain.
"Responsibility" is definitely a multifaceted and multilayered
concept and there are different levels of involvement in terrorist activities.
Anti-terrorism measures must take into consideration the extent of
involvement that the individuals targeted by these measures are suspected
of. People who finance terrorist organizations, for instance, may be subject
to the freezing of their assets without recourse to the criminal law route, but
not to detention.
The third feature of national security decisions touches on the
characteristics of the people affected by anti-terrorism measures, who are
often members of groups less likely to be harmed by terrorism on the one
hand, and more likely to be targeted by government action on the other.
Usually, terrorists are not members of the societies they attack. This is
obviously the case if they are citizens of another state, but also when they
live in closed and distinctive communities less likely to be active in sites
targeted for attack. Burdening them with the costs of anti-terrorism may
balance their initial inclination to refrain from interest in the effects of the
attacks. The opposite is also true: victims of terrorism are often targeted
because of their group affiliation, and the choice of victims is not wholly
random. Accordingly, expecting governments to adhere to considerations of
global justice33 is less realistic when the initial threat is not equally allocated
but rather affected by nationality.
This factor also points to the relative exclusion from the political
process of those liable to be targeted by anti-terrorism policies, usually
because they are foreigners or members of insular minority communities.
These characteristics allow for an interesting juxtaposition with the analysis
of public choice theory concerning the circumstances requiring
governments to choose between allocating costs to a small group or to the
public at large. According to standard public choice analysis, one should
fear that when decision makers face a choice between inflicting harsh
burdens on a small group and distributing the cost of regulation among all
See Thomas W. Pogge, GlobalJustice, in GLOBALJUSTICL 3-5 (Thomas W. Pogge, ed.,
2001).
297
HarvardNational Security Journal/ Vol. 3
citizens, they will tend to prioritize the interests of the small interest group,
which is usually better organized and more effective at promoting its own
interests.3 4 In contrast, in the typical national security scenario, we should
be guided by the opposite fear-that the general public will succeed in
shifting the burden of national security to "discrete and insular
minorities,"3 and even more so to people who, very often, are not
represented in the political process because they are not citizens and, at
times, are even illegal residents. Often, decisions in this area are not made
behind a "veil of ignorance," to use Rawls' famous metaphor. 36 The public
and its representatives can identify those who would bear the burden of
various anti-terrorism measures. Accordingly, in the context of national
security, the concern should be the opposite to the one stressed by standard
public choice analysis-that moral panic and the lack of any "veil of
ignorance" regarding those who may be made to bear the burdens of
national security measures will disproportionately tilt the balance of
decisions in the other direction.
The fourth feature of national security decisions is that the burdens
attached to some of them (usually those causing irreversible harm) are
difficult or even impossible to spread. In most regulatory contexts,
considering the possibility of spreading costs is relevant when at stake is a
policy choice in which a small group of actors is expected to carry a heavy
burden. When the costs can be spread to others, the problem of allocating
the burden to a limited group is less pressing. Yet, in contrast to polluters,
for instance, who can spread the cost of the pollution taxes imposed on them
to their contractors and consumers, suspected terrorists against whom
harsher anti-terrorism measures are used are burdened by government acts
that fall entirely on their shoulders. How can an individual subjected to
administrative detention spread the costs of being deprived of liberty?
3See generally DANILL A. FARBER & PHILIP P. FRICKLY, LAW AND PUBLIC CHOICE: A
CRITICAL INTRODUCTION 12-37 (1991).
- This well-known phrase is taken from the opinion ofJustice Stone in United States v.
Carolene Products Company., 304 U.S. 144, 152 n.4 (1938).
6 RAWLS, supra note 18., at 136 ("The idea of the original position is to set up a fair
procedure so that any principle agreed to will be just. The aim is to use the notion of pure
procedural justice as a basis of theory. Somehow we must nullify the effects of specific
contingencies which put men at odds and tempt them to exploit social and natural
circumstances to their own advantage. Now in order to do this I assume that the parties are
situated behind a veil of ignorance. They do not know how the various alternatives will
affect their own particular case and they are obliged to evaluate principles solely on the
basis of general considerations.").
2012 /Distributihe Justice in National Security Law
298
Compensation does not seem to fully balance this burden (and even if
allocated ex post, it cannot be considered a full ex ante justification for
imposing burdens that have non-monetary consequences).37
Indeed, as noted, some of the four factors discussed above are also
found in other regulatory contexts. The uniqueness of the national security
context is that it combines all four. When the government must choose
between instituting a maximum driving age of sixty-five or reducing speed
limits for all drivers, it must compare different types of regulatory burdens
allocated to different groups. Aging drivers, however, cannot be considered
to carry any moral blame, and they are also effectively represented in the
domestic political process. Another oft-discussed example concerns
regulation in environmental matters. When a government wishing to keep
water resources clean is forced to choose between allocating costs to
factories that operate in the area by creating a new tax imposed solely on
them (given findings that they are at least partially responsible for water
pollution) or implementing a tax that will affect the population at large
(through the regular tax system), it chooses between "responsible" and
"innocent" cost-bearers. The cost, however, is of a similar kind (monetary)
and more significantly, if imposed on the factories, they can spread the costs
to others. Factories, moreover, can voice their arguments and influence the
political arena. The special nature of the distributive justice challenge in the
context of national security has a further aspect. Contrary to other areas of
regulation, government action for the promotion of national security usually
involves the regulation of private parties as well as the provision of
government services given that national security, as noted, is also a public
good.38
III. The Institutional Aspects of the Distributive Justice Challenge in
National Security Law
Distributive justice, then, should be regarded a central consideration
affecting regulatory choices made in the name of national security. Due to
the institutional features of policy formulation in this area, however, the
7 For an argument in support of awarding compensation to those subjected to detentions
in the area of national security, see, e.g., Eugene Kontorovich., Liability Rulesfor Constitutional
Rigits: 7ie Case ofMass Detentions, 56 STAN. L. RLV. 755 (2004).
8 For example, security checks at airports can be performed either by private agents who
follow policies and regulations promulgated by the government, or directly by government
agents.
299
HarvardNational Security Journal/ Vol. 3
likelihood of addressing this aspect of national security law in the courts or
in the public discourse is low.
When courts review the legality and reasonableness of anti-terrorism
measures, their focus is limited to the issues placed before them. Whether
these involve the use of physical measures in interrogations, control orders,
or wiretapping, the courts assess each one on an individual basis and, as a
structural matter, fail to compare the measures to other possible antiterrorism options. This is also generally the case even in jurisdictions that
use the "proportionality" doctrine, which mandates courts to question
whether no alternative methods could have furthered the stated government
goal with less infringement of human rights. 39 This result reflects both the
extent of deference that is widespread in the context of national security and
the limitations of the proportionality test that focuses on immediate
alternatives to the decision at hand rather than on different policy
approaches, such as preferring the use of large-scale surveillance over harsh
interrogations.
But even if judges were willing to expand their review of national
security decisions so as to include consideration of completely different
alternative measures, they would probably face the hurdle of the
government's advantage of expertise in this area. This expertise may enable
it to manipulate the evaluation of alternatives by arguing that they are not
equally effective.
Similar obstacles erode the chances of robust public scrutiny of
national security policies. Secrecy considerations keep most of the relevant
information required to evaluate regulatory decisions in the area of national
security out of the public domain. As a result, they bar effective discussion of
whether the government should have approached security concerns in an
entirely different manner and, for instance, not only use less harmful
measures of interrogation but also abandon them altogether and shift its
efforts to surveillance.
Note also that judicial and public review mechanisms are not likely
to possess the full picture about the distributive implications of each decision
they evaluate, and pressures may lead them to decide on changes that do
not fully consider all these implications. For example, human rights
* Alec Stone Sweet &Jud Mathews, ProportionalityBalancing and Global Constitutionalism, 47
CoLUA.J. TRANSNAT'L L. 72, 75-76 (2008).
2012 /Distributihe
Justice in NationalSecurity Law
300
lobbyists may succeed in shortening the time allowed for interrogating
terrorist suspects, leading to the broader use of other surveillance measures
that affect the privacy of the population at large. Similarly, litigation against
security measures promoted by human rights organizations may not always
lead to a better balance between human rights and security. Rather, it may
shift the cost of national security to other individuals, who will incur the
burden of the alternative measures adopted if the government is forced to
abandon its previous policy. The judicial process is not well-adapted to deal
with systemic changes brought about by litigation. At best, it leaves the door
open to the future review of regulatory decisions shaped by earlier judicial
precedents.
The importance of parliamentary review and of internal oversight
mechanisms of national security agencies, therefore, can hardly be
overestimated. Unlike the courts and the public at large, these institutions
do enjoy broad access to relevant information. The emphasis on internal
review for the purpose of evaluating the distributive effects of national
security decisions goes hand in hand with a growing understanding that
effective oversight of executive decisions must resort to these mechanisms in
various areas of government action. 40
IV. The Distributive Justice Challenge of National Security Law
in Real Life Case-Studies
The need for choosing between placing large burdens on the
members of a small group as opposed to allocating lesser burdens to society
at large has not been directly discussed in the literature on national security.
Some of the analyses touching specifically on anti-terrorism dilemmas,
however, do imply it, as the following discussion points out.
A classic example concerns the debate surrounding the use of
particular methods in the interrogations of terrorism suspects. Some
arguments against the use of such measures emphasize that their use is not
vital given the availability of alternative methods, such as effective
4
See, e.g., Neal K. Katyal., Internal Separation ofPowers: Checking Today's Most DangerousBranch
from Within, 115 YALE L.J. 2314 (2005) (discussing internal restraints on executive branch
action). In the area of national security, the distributive justice implications of various
policies must be part of the ongoing dialogue between the agencies that participate in the
formulation of these policies. Details would obviously vary in different systems. Israel's
National Security Council., for instance, is an advisory body to the Prime Minister and the
government and, as such, is potentially well situated for the evaluation of these questions.
30 1
HarvardNational Security Journal/ Vol. 3
intelligence work, which may lead to the uncovering of the same
information. 4 ' But this alternative carries its own costs since it involves the
large-scale use of wiretapping and other privacy infringements affecting
large numbers of people. This particular example may also be considered
problematic from a deontological perspective given the absolute prohibition
of torture,4 2 but we can still study it inasmuch as it does not address torture
but rather a choice between harsh interrogations that fall short of the use of
torture and pervasive wiretapping. Another version of this dilemma,
involving less severe measures, addresses the choice between issuing control
orders against individuals involved to some extent in terrorist activities (in a
manner that will enable the gathering of information about their activities
more effectively and decrease their ability to engage in dangerous acts),4 3
and the indiscriminate collection of information about the population at
large (e.g., by placing cameras in public places and other measures). The
latter entails more widespread privacy infringements, but is less harmful to
any particular individual involved.
Another, more concrete, example of a similar regulatory choice
taken from the Israeli context is the building of a security barrier to guard
civilian areas vis-a-vis the targeted killing of key terrorists. The Israeli
government has justified its building of the barrier on the grounds that a
physical obstacle is necessary between Israel and the occupied territories so
as to prevent the entry of suicide bombers into Israel. Without delving into
the controversial discussion of whether the building of the barrier was also
41 For the need to concentrate on intelligence tools in the fight against terrorism, see
PHILLIP B. HEYMANN, TLRRORISM, FRLLDOM, AND SLCURITY: WINNING WITHOUT WAR
(2003). For an evaluation of Heymann's proposals and their risks, see Ronald D. Lee &
Paul M. Schwartz, Beyond the "War" on Terrorisn: Towards the New Intelligence Network, 103
MICH. L. REv. 1446 (2005). For additional discussion on the focus on intelligence as a
central anti-terrorism measure, see also RICHARD A. POSNLR, COUNTLRING TLRRORISM:
BLURRED FOCUS, HALTING STLPS (2007); RICHARD A. POSNER, PRLVLNTING SURPRISE
ATTACKS: INTELLIGLNCL REFORM IN THL WAKE OF 9/11 (2005); RICHARD A. POSNLR,
UNCERTAIN SHILED: THE U.S. INTELLIGENCL SYSTEM IN THL THROLS O REFORM
(2006).
42 This approach was adopted by the International Convention Against Torture and Other
Cruel, Inhuman or Degrading Treatment or Punishment., Dec. 10, 1984, S. Treaty Doc.
100-20, 1465 U.N.T.S. 85, available at http://www2.ohchr.org/english/law/cat.htm.
1< This measure, which involves significant infringement of liberty but falls short of
detention, has been used in British law since 2005. See Clive Walker, Keeping Controlof
Terroists Without Losing Control ofConstitutionalisn, 59 STAN. L. REv. 1395, 1402 (2007).
2012 /Distributihe
Justice in NationalSecurity Law
302
motivated by political purposes,4 4 the decision to build the barrier may be
compared to other anti-terrorism measures directed only against individuals
claimed to be involved in terrorism, of which targeted killings are the most
extreme example.> For the purposes of this Article, the important
connection is the following: the security barrier and the targeted killings are
two measures that contribute to the safety of Israeli citizens and should be
understood as responses to the same problem. Whereas one entails a largescale infringement of rights, the other is a far more harmful act directed at a
much smaller group of people. In some circumstances, at least theoretically,
the same level of security might be achieved through the building of physical
barriers affecting large populations or through the deadly targeting of active
terrorists. 46
Besides these relatively detailed examples, note that other legal
initiatives in the area of national security also have similar features in that
they require a choice between preventive measures that are universally
applicable and others that are more focused. For instance, in the context of
laws on terror financing, some measures such as mandates to disclose the
sources of money transfers are aimed at the population at large, while other,
more severe, legal actions, such as the freezing of personal bank accounts,
only target individuals suspected of supporting terrorism. 4 7 Although these
regulatory schemes often complement one another, they may also involve
choices, such as whether to broaden universally applicable regulation or
place additional burdens only on suspects. Even targeted measures, such as
wiretapping the Internet communications of suspected terrorists, can have
implications for the population at large because surveillance is made
possible by laws mandating the use of certain computer technologies.4 8
See generally Daphne Barak-Erez., The Sectfity Ban ier: Between InternationalLaw, Constitutional
Law and Domestic JudicialRev iew, 4 INT'LJ. CONST. L. 540 (2006).
5 See HCJ 769/02 Public Committee Against Torture in Israel v. Government of Israel
62(1) IsrSC 507 [2006] (Isr.).
46 In practice., Israel has used both measures, but in principle, it is plausible that the use of
targeted killings may be narrowed or broadened according to the use of other preventive
measures such as the security barrier.
47 SeeJoined Cases C-402P & C-415/05P, Yassin Abdullah Kadi &Al Barakaat Int'l
Foundation v Council of the European Union & Comm'n of the European Communities,
2008 E.C.R. I-635 1.
48 See Charlie Savage, U.S. Tfies to Make It Easie, to Wiretap the Internet, N.Y. TIMES, Sept. 27,
2010, at Al, available at
http://www.nytimes.com/2010/09/27/us/27wiretap.html?pagewanted=all.
303
HarvardNational Security Journal/ Vol. 3
These examples should be distinguished from others that may
involve distributive choices but do not involve all the features of the more
difficult cases this discussion focuses on. When facing the decision of
whether to use racial profiling as an anti-terrorism measure, a government
must choose between allocating the costs to a small group (such as all people
with certain ethnic features) or to the general public (or an unidentified large
segment of it, such as all those who visit airports). 49 In this example,
however, those who belong to the small group are morally no more
blameworthy than the population at large.50 This example, therefore,
exceeds the scope of the current analysis, which does not promote the
allocation of burdens on the basis of immutable or arbitrary individual
traits. 1 For similar reasons, the examples discussed above are also different
from the moral problem known as the "Sheriffs Dilemma," which involves
the saving of the innocent majority through the sacrifice of innocent
individuals by turning them over to a murderous mob.52
See FREDERICK F. SCHAUER, PROFILES, PROBABILITIES AND STEREOTYPES 190 (2003)
(discussing the possibility that profiling may be avoided by screening all air-travellers).
* Even in the classic cases discussed above, as noted, there could be varying degrees of
evidential uncertainty around the responsibility of individuals targeted for the application of
selective anti-terrorism measures. In racial profiling, however, the individuals subjected to
additional burdens are, by definition, innocent.
' See RONALD D WORKIN, SOVEREIGN VIRTUE: THE THEORY AND PRACTICE OF
EQUALITY 287 (2000) ("individuals should be relieved of consequential responsibility for
those unfortunate features of their situation that are brute bad luck, but not from those that
should be seen as flowing from their own choices"). Note, however, that renewed support
for various forms of profiling is emerging with the relatively new recourse to universal uses
of body-scans in airports. See, e.g., Eugene Robinson, TSA Outcy is Really a Callfor Profiling,
VASH. POST (Nov. 23, 2010), available at http:/ /www.washingtonpost.com/wpdyn/content/article/2010/11/22/AR2010112204387.html; Susan Stellin, Support Growsfor
Tiered Risk System atAirports, N.Y. TIMES, Feb. 7 2011, at B6, available at
http://www.nytimes.com/2011/02/08/business/08security.html.
2 "In a frontier town three children have been abducted, sexually assaulted, tortured and
murdered. There is an enormous public demand that the local sheriff find the murderer. As
time goes on and no one is arrested, public fear increases, unrest grows and confidence in
the forces of law and order diminishes. A man is arrested, and such is the circumstantial
evidence against him that it is widely believed that the real murderer has been found. It
becomes clear to the sheriff that the man he has arrested is innocent and ought to be
released, but a lynch mob has gathered and is threatening to tear down the jailhouse unless
the suspect is tried and executed or handed over. There is no immediate possibility of a fair
trial, but it looks to the sheriff as though serious public disorder and considerable damage
and injury are likely if he tries to resist the demands of the lynch mob. Should he execute or
hand over to the mob a man he knows to be innocent?" GORDON GRAHAM, EIGHT
THEORIES O1 ETHICS 145 (2004). As in the example of ethnic profiling, the arrested man is
not more blameworthy than the rest of the town people.
4
2012 /Distributive
Justice in NationalSecurity Law
304
V. Thinking about Distributive Justice in National Security Law
The discussion so far has described how distributive concerns affect
national security decisions. Understanding these connections can now serve
as a basis for discussing the relative advantages and disadvantages of each of
the following alternatives: taking measures that may involve significant
human rights infringements but affect only select individuals vis-A-vis taking
measures that involve lesser harms which are universally applied.
A. Responsibility Considerationsvs. Public Choice Considerations
Analyzing the alternatives solely through a moral responsibility
prism could lead to the conclusion that allocating national security burdens
to those who engage in terrorist activities is a fair solution. Although each
individual thereby incurs a relatively high cost, this solution is preferable to
the allocation of risks, even if less burdensome ones, to innocent people.
And yet, proposing answers in this area without considering the
ramifications of political pressures ignores the danger of overusing harsh
specifically targeted measures. When national security is concerned, the
public can be expected to urge the use of harsh measures against terrorist
suspects on the assumption that this will affect only "foreigners" and
"others." The participants in the debate therefore formulate their
preferences without the "veil of ignorance" that should be used in
evaluating policy decisions, which could result in the overuse of
disproportionate measures aimed at those perceived as "others." Given this
bias, advocating the use of anti-terrorism measures that target the public
more generally appears advisable because such measures will be debated
much more seriously. 3 A further aspect of this consideration evoking grave
concern is that the pressure of public panic could lead the government to err
in the direction of over-inclusiveness when defining the specific group
targeted for anti-terrorism measures so that, for instance, it will include not
At the same time, note that even investigation and law enforcement measures practiced
on a universal basis may entail hidden distributive consequences. For example, it has been
argued that Fourth Amendment protections against searches tend to benefit more wealthy
populations because they apply to houses and cars and much less, if at all, to people who
conduct their daily activities in the streets. See WilliamJ. Stuntz, The Distribution ofFouth
Amendnent PriVay, 67 GEO. VASH. L. RLV. 1265, 1270 (1999).
HarvardNational Security Journal/ Vol. 3
305
only participants
members. 4
in terrorist-related
activities
but also their family
B. The Implications of Uncertainty and PossibleMistakes
Another concern evoked by the imposition of national security
measures focusing on suspected terrorists rather than on the population at
large touches on the factor of uncertainty. By definition, suspected terrorists
are still only suspects, which implies a possibility of mistakes. Sensitivity to
mistakes should increase proportionate to the severity of the anti-terrorism
measure used, reaching a peak when targeted killings, detentions, harsh
interrogations, and the like are involved. Therefore, the possibility of
mistake is a significant consideration against the use of harsh measures
targeting individuals on the basis of evidence collected through
administrative means. Lack of certainty does not necessarily imply that such
measures should be avoided altogether, but may lead to the prioritization of
other means of prevention, if available.
C. ChancesforjudicialReview
Another factor to consider when assessing alternative anti-terrorism
measures is the probability that they will be subjected to judicial review, the
assumption being that anti-terrorism measures that have a higher chance of
being reviewed by courts should be considered better, in the sense that they
will not go unnoticed and remain unbalanced. From this perspective, note
that harsher measures directed at specific individuals are more likely to be
reviewed by courts than more dispersed infringements, both because of
lesser incentives for individuals to initiate proceedings and because of
potential barriers in systems requiring individuals to be personally affected
so as to establish standing.55 The use of policing and investigation measures
against large groups or against the population at large (such as roadblocks,
cameras, and so forth) without a basis of suspicion have been litigated in
54 An even more inclusive approach may lead to the application of certain policies to all the
residents of an area in which terrorist activities take place. See supra note 30.
II This is the U.S. approach. For example., the challenge to the warrantless wiretapping
program that began as part of the new anti-terrorism measures adopted after September 11
was not decided on its merits. The Sixth Circuit found that the plaintiffs, a collection of
organizations and individuals who were involved in national security litigation, had failed to
substantiate sufficient standing because they could not show with certainty that they had
been or would be subject to any wiretapping under the policy. Am. Civil Liberties Union v.
Nat'l Sec. Admin., 493 F.3d 644 (2007), cert denied, 552 U.S. 1179 (2008).
2012 /Distributihe
Justice in NationalSecurity Law
306
U.S. courts relying on the Fourth Amendment to the U.S. Constitution,
which states that "no warrants shall issue, but upon probable cause."5 6 In
most cases, however, these challenges have been dismissed, with courts
holding that these measures did not infringe the constitutional guarantee as
long as they included relatively light burdens and were not applied in a
discriminatory manner. 7
Beyond these considerations, and leaving aside the question of the
better choice-shifting the burden of infringement to the few or to the
many-the distributive justice perspective on national security policies is
important as a platform for evaluating and criticizing current policies. When
debating a national security policy, criticizing the extent to which it infringes
human rights will not suffice. Another vital element is to evaluate it vis-a-vis
alternatives that achieve the same national security goal by infringing on the
rights of different people in different ways.
Conclusion
This Article argues that the regulation of national security matters is
greatly influenced by distributive choices that have not yet been fully
acknowledged. Recognizing the relevance of distributive justice to national
security law has several consequences. First, it bears implications for the
theory of distributive justice itself, which should increasingly turn its
attention to the distribution of burdens in the formulation of policies.
Second, distributive justice concerns bear significant implications for the
institutional study of national security law. Given the relative inability of
courts to consider alternatives beyond the immediate issue before them, the
responsibility of executive and legislative decision makers in this area
becomes even more significant. Third, and more concretely, the Article
- U.S. CONST. amend IV.
For an assessment of the case law on these relatively-spread policing measures, see
Christopher Slobogin., GovernmentDragnets, 73 L. & CONTEMP. PROB. 107 (2010). Tracey
Meares and Dan Kahan have similarly advocated the possibility of enabling inner-city
communities to compromise the rights of all their members in a manner that would enable
searches and privacy infringements against everyone (with no specific cause) in the name of
promoting the overall security of these communities and of all the individuals living in
them. This analysis took into consideration "burden-sharing" by all community-members,
in a way that did not single out a particular group. According to their argument, this
burden-sharing relieves concerns of discrimination and singling-out, and should therefore
be immune from judicial review. See Tracey L. Meares & Dan Kagan, When Rights are
Wrong: The Paradoxof Unwanted Rights, in URGENT TRILS: POLICING AND RIGHTS IN INNERCITIEIS COMMUNITIES 23-24 Joshua Cohen &Joel Rogers, eds., 1999).
5
307
HarvardNational Security Journal/ Vol. 3
begins to define the criteria to be taken into account when choosing
between alternative measures that promote national security. From a moral
perspective that highlights the responsibility of individuals for their choices,
anti-terrorism measures focused on individuals who are involved (or
partially involved) in terrorism-related activities appear to hold an
advantage. Preference for policies that rely on harsh and targeted measures
may also be inspired by the pragmatic consideration that such policies are
more likely to come up for judicial review due to their relative severity.
Other considerations, however, lean in the opposite direction. Since faultrelated measures usually target individuals outside the polity or belonging to
discrete and insular communities, over reliance on harsh measures against
them becomes a real danger. From this perspective, a policy inspired by
public choice understandings would see advantages in anti-terrorism
measures applied on a universal basis. In addition, these measures have the
advantage of being less likely to cause grave injustice due to mistakes in
ascertaining the fault of specific individuals.
2012 /Distributive
Justice in NationalSecurity Law
308